A Novel, Scalable Social Media-Based Intervention ("Warna-Warni Waktu") to Reduce Body Dissatisfaction Among Young Indonesian Women: Protocol for ...
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
JMIR RESEARCH PROTOCOLS Garbett et al
Protocol
A Novel, Scalable Social Media–Based Intervention (“Warna-Warni
Waktu”) to Reduce Body Dissatisfaction Among Young Indonesian
Women: Protocol for a Parallel Randomized Controlled Trial
Kirsty May Garbett1, BA, MSc; Nadia Craddock1, BSc, EdM, PhD; Sharon Haywood1, BSc, MSc; Kholisah Nasution2,
MIPH, MD; Paul White3, BSc, MSc, PhD; L Ayu Saraswati4, MA, SIP, PhD; Bernie Endyarni Medise2, MPH, MD,
PhD; Girl Effect; Percolate Galactic; Phillippa C Diedrichs1, BSc, PhD; Heidi Williamson1, BSc, MSc, DHealth
1
Centre for Appearance Research, University of the West of England, Bristol, United Kingdom
2
Faculty of Medicine, Universitas Indonesia, Jakarta, Indonesia
3
Applied Statistics Group, University of the West of England, Bristol, United Kingdom
4
Department of Women, Gender, and Sexuality Studies, University of Hawaii, Manoa, HI, United States
Corresponding Author:
Kirsty May Garbett, BA, MSc
Centre for Appearance Research
University of the West of England
Frenchay Campus
Coldharbour Lane
Bristol, BS16 1QY
United Kingdom
Phone: 44 1173282911
Email: Kirsty.garbett@uwe.ac.uk
Abstract
Background: Despite the prevalence of body dissatisfaction among young Indonesian women and its consequential negative
impacts, there are currently no evidence-based, culturally appropriate interventions to tackle this issue. Therefore, there is a need
to develop scalable, cost-effective, and accessible interventions to improve body image among this population.
Objective: This paper describes the study protocol of a parallel randomized controlled trial to evaluate the effectiveness of
Warna-Warni Waktu, a social media–based intervention that aims to reduce state and trait body dissatisfaction and improve mood
among young Indonesian women aged 15-19 years.
Methods: The trial will take place online. Approximately 1800 young women from 10 cities in Indonesia, evenly split across
the ages of 15-19 years, will be recruited via a local research agency’s established research panel. Participants will be randomly
allocated to the intervention condition or a waitlist control condition. The intervention consists of six 5-minute videos, with each
video supplemented with up to five brief interactive activities. The videos (and associated activities) will be delivered at a rate
of one per day across 6 days. All participants will complete three self-report assessments: at baseline (Day 1), 1 day following
the intervention (Day 9), and 1 month following the intervention (Day 36). The primary outcome will be change in trait body
dissatisfaction. Secondary outcomes include change in internalization of appearance ideals, trait mood, and skin shade satisfaction.
Intervention effectiveness on these outcomes will be analyzed using linear mixed models by a statistician blinded to the randomized
condition. Intervention participants will also complete state measures of body satisfaction and mood before and after watching
each video to assess the immediate impact of each video. This secondary analysis of state measures will be conducted at the
within-group level.
Results: Recruitment began in October 2021, with baseline assessments underway shortly thereafter. The results of the study
will be submitted for publication in 2022.
Conclusions: This is the first study to evaluate an eHealth intervention aimed at reducing body dissatisfaction among young
Indonesian women. If effective, the intervention will be disseminated to over half a million young women in Indonesia via
Facebook, Instagram, and YouTube.
Trial Registration: ClinicalTrials.gov NCT05023213; https://clinicaltrials.gov/ct2/show/NCT05023213
International Registered Report Identifier (IRRID): PRR1-10.2196/33596
https://www.researchprotocols.org/2022/1/e33596 JMIR Res Protoc 2022 | vol. 11 | iss. 1 | e33596 | p. 1
(page number not for citation purposes)
XSL• FO
RenderX
JMIR RESEARCH PROTOCOLS Garbett et al
(JMIR Res Protoc 2022;11(1):e33596) doi: 10.2196/33596
KEYWORDS
body image; body dissatisfaction; Indonesia; adolescent; mental health; randomized controlled trial; study protocol; eHealth
intervention; Southeast Asia; young adult; teenager; women; social media; intervention; image; protocol; mood; satisfaction
providing effective mental health treatment [29,30]. Although
Introduction Indonesia has seen a modest increase in mental health
Body dissatisfaction, defined as the subjective experience of interventions for adolescents in school settings in recent years
negative thoughts toward one’s own body [1], is a growing [31-33], evidence suggests that implementation is not
concern among young people globally [2]. Young women are commonplace [34], thus highlighting the need for the
disproportionately affected compared to young men [3-5]. dissemination of educational mental health content outside the
Although research in Asia is currently limited, cross-cultural school system. Digital interventions present a solution to
research suggests that the prevalence of body dissatisfaction circumvent barriers to dissemination, and have been shown to
among young people in some Asian countries such as Malaysia, have similar effectiveness as face-to-face psychotherapeutic
China, and Japan is similar to, if not greater than, that of interventions [35]. Digital interventions are relatively low in
English-speaking countries [6-8]. cost, easily accessible, and universally available, thereby
meeting three key criteria for overcoming LMIC-specific
Body dissatisfaction is not benign. Extensive research has challenges [36,37]. Further, private, remotely accessible mental
established that body dissatisfaction is associated with numerous health interventions have been shown to increase the likelihood
adverse health outcomes. Longitudinal research has found body of engagement in help-seeking due to the reduced fear of stigma
dissatisfaction to predict eating disorders [9,10], depressive [38].
mood and low self-esteem [11], less engagement in exercise
[10], and increased risky health behaviors such as drug use and eHealth interventions have the potential to reach many young
smoking [12]. Research also indicates associations between people in Indonesia; according to the Indonesian Internet
body dissatisfaction and the desire for cosmetic surgery [13,14], Providers Association, over 90% of adolescents aged 15-19
and the avoidance of everyday activities such as participating years have access to the internet across the country [39].
in school activities or attending classes [15,16]. Although the Moreover, research indicates that young people are already
bulk of this research has been concentrated within using the internet to seek information [40,41]. Thus, it is perhaps
English-speaking populations, similar associations are emerging unsurprising that eHealth interventions for mental health have
globally, including across Asia [6,17-20]. shown preliminary acceptability among young people in
Indonesia. For example, eHealth interventions for depression
Indonesia is an upper-middle-income country in Southeast Asia prevention and/or treatment have been well received, with young
and is the world’s fourth most populous country [21]. In the people showing a willingness to engage with such interventions
Indonesian context, research shows that more than half of young [42,43].
women experience at least some dissatisfaction with their
appearance (unpublished data compiled by the authors KG, NC, Social media offer unprecedented capabilities to disseminate
LS, and PD, 2021) [22]. Body dissatisfaction among young mental health interventions cost-effectively and at scale [44],
Indonesian women is linked with disordered eating behaviors and may be particularly popular with young people. Social
[23], including food restriction and avoidance, emotional eating, media interventions afford unique opportunities to overcome
and excessive exercise (unpublished data by authors KG, NC, barriers such as cost, geographic distance, and stigma, as they
LS, and PD, 2021). Further, young Indonesian women report allow for a certain degree of privacy and anonymity. Emerging
body dissatisfaction specifically with regard to their skin shade research suggests that using social media in Indonesia as a
[24,25], due in part to the dominant pan-Asian ideal prevalent vehicle for eHealth interventions shows promise among young
across Asia, which emphasizes light skin shades [26]. In other people [45,46]. Thus, the combination of the rise in eHealth
populations (eg, India), skin shade dissatisfaction has been initiatives for adolescents in LMICs [47] and the notable uptake
associated with the use of potentially harmful skin fairness in social media usage among young people in Indonesia in recent
products [27], lower self-esteem [28], and, unsurprisingly, years [48,49] have created a ripe environment for the
general body image concerns [28]. As such, ameliorating body development of a social media–based intervention to address
dissatisfaction, including skin shade dissatisfaction, among body dissatisfaction among young Indonesian women.
young women in Indonesia is required. Yet, no published Furthermore, research consistently shows the social media
evaluations of interventions targeting body dissatisfaction in environment in general to be problematic for young people’s
Indonesia exist. body image [50,51], due in large part to the objectification and
idealization of women’s bodies. As such, hosting an intervention
The creation and dissemination of mental health prevention to reduce body dissatisfaction on social media may be
efforts and interventions across low- and middle-income additionally beneficial in disrupting the harmful effects these
countries (LMICs), including Indonesia, face cultural-specific sites have been shown to have. Increasingly, the potential for
obstacles, primarily social stigma connected to mental health positive content on social media is being explored, which has
concerns, coupled with a lack of mental health professionals
and the limited capacity of general health professionals in
https://www.researchprotocols.org/2022/1/e33596 JMIR Res Protoc 2022 | vol. 11 | iss. 1 | e33596 | p. 2
(page number not for citation purposes)
XSL• FO
RenderXJMIR RESEARCH PROTOCOLS Garbett et al
highlighted the benefits of some types of content on mood and The videos target three sociocultural influences of body
body satisfaction [52,53]. dissatisfaction, based on the Tripartite Influence Model of body
dissatisfaction [58], namely, the media, friends, and family. The
This paper outlines the development and protocol for the
Tripartite Influence Model postulates that body dissatisfaction
evaluation of the first social media–based intervention to target
increases via the impact that these sociocultural influences have
body dissatisfaction among young women in Indonesia, named
on two psychological processes: internalization of appearance
Warna-Warni Waktu (English translation: Colorful Time
ideals and social comparisons. These two psychological
Travel). The intervention was developed by the academic
processes are also addressed directly in the videos by providing
authors of this paper in collaboration with Girl Effect, an
media literacy education and examining the consequences of
international nonprofit organization that builds media content
making appearance-based social comparisons. As such, we
aiming to arm girls with the skills to make positive choices and
anticipate that the intervention will reduce body dissatisfaction
changes in their lives during the critical years of adolescence;
through diminishing an individual’s perceived appearance
the Dove Self-Esteem Project, the social mission for Unilever’s
pressure from the media (including social media), friends, and
personal care brand, Dove; Percolate Galactic, an
family members, which in turn will reduce their internalization
Indonesian-based creative agency that specializes in marketing
of appearance ideals and the likelihood of making social
for youth; and young Indonesian women. The intervention
comparisons.
consists of a series of six videos, each approximately 5 minutes
long. To disseminate the intervention, the videos will be The potential impact of the videos is further bolstered through
sequentially delivered to young women in Indonesia through supplementary activities, which again will be disseminated to
targeted social media marketing on Facebook and Instagram, young Indonesian women via sequential social media marketing.
in addition to being made freely available on YouTube. On Each video is accompanied by interactive activities, which aim
Facebook and Instagram, the videos will appear on young to reinforce the lessons learned in each video. Research
Indonesian women’s feeds, consistent with other social media consistently shows that elements of active learning (ie, taking
advertisements. The series tells the fictitious story of a young control of one’s own learning through metacognitive
woman named Putri who learns strategies to resist appearance sense-making, self-assessment, and reflection [59]) results in
pressures across adolescence and young adulthood through the deeper learning [60] and higher engagement levels [61]. Further,
help of animated time travelers, who are on a quest to save the eliciting cognitive dissonance has been shown to be a key
world from appearance-related pressures. The intervention is change technique in reducing body dissatisfaction [54]. As such,
based upon mounting evidence that psychoeducation, in active learning activities based upon cognitive dissonance were
particular discussing the nature, causes, and consequences of built into the intervention to be delivered between each video.
body dissatisfaction, is an effective change technique to reduce These include activities such as story completion, self-reflection,
body dissatisfaction [54]. Further, the intervention’s videos writing challenges, and word searches. Details of each activity,
model behaviors to reduce appearance pressure and teach media along with a breakdown of each video’s content, are outlined
literacy skills, two further change techniques that have shown in Textbox 1.
efficacy in reducing body dissatisfaction in previous studies
[55-57].
https://www.researchprotocols.org/2022/1/e33596 JMIR Res Protoc 2022 | vol. 11 | iss. 1 | e33596 | p. 3
(page number not for citation purposes)
XSL• FO
RenderXJMIR RESEARCH PROTOCOLS Garbett et al
Textbox 1. Intervention summary.
Video One: Time to Turn Back Time
Key messages:
• People face pressure to look a certain way.
• Pressure comes from the media, as well as from those around us.
• Appearance pressures are associated with body dissatisfaction and can hold us back from living a fulfilling life.
• Strategies can be learned to resist and challenge appearance pressures.
Reinforcer activities:
• Review definitions of key terms in five short videos. Using some of the key terms, explain why the time travelers want Indonesian girls to feel
confident about their bodies.
Video Two: That’s Fake! (targeting social media)
Key Messages:
• The media often portray just one narrow appearance ideal.
• Images of people in the media are often edited to make the person look more like the appearance ideal.
• The media set an unrealistic appearance standard in order to sell us products.
• We can curate our own media environment to reduce the appearance pressures that we face.
Reinforcer activities:
• Describe your experience of coming across advertisements that promote unrealistic beauty products to help others become more aware.
• Share the transformation video clip seen in this episode on social media, and explain why it is important your friends and followers watch it.
• An edited and unedited image is provided. Identify and list all the edits made to the image.
• Watch the video Putri saw advertising skin-lightening cream, and critically examine its messages about the lifestyle the advert is trying to sell.
Share your experience buying (or thinking about buying) a product that you thought would improve your popularity or lifestyle.
• Your Own Words activity: Write about why skin-lightening products are problematic.
Video Three: C’mon, Break the Chain of Comparisons (targeting appearance-based comparisons)
Key messages:
• Appearance-based comparisons are common.
• Engaging in appearance-based comparisons is unhelpful and damaging to body image.
• Focusing on what your body can do, rather than how it looks, is a more helpful way to think about your body.
Reinforcer activities:
• Write a sentence or two on why you appreciate your friends, describing things that have nothing to do with appearance.
• Complete the comic strip: In the context of appearance-based comparisons among friends, explain why appearance-based comparisons are not
helpful or necessary.
• Your Own Words activity: Write about all the things you love about your friends that have nothing to do with appearance.
Video Four: Stand up to Appearance-Based Comments (targeting appearance-based teasing)
Key messages:
• Comments from friends and family about our appearance can be hurtful, even if well-meaning.
• Challenging such comments in a nonconfrontational way can prevent future comments from family members, alleviating appearance pressures.
Reinforcer activities:
• Write a short response to a negative appearance-based comment received from a friend or family member.
• Word search: Find 10 hidden example phrases of how to respond to appearance-based comments.
• Complete the comic strip: How to respond to boys teasing girls about their appearance.
• Your Own Words activity: Write about how to stand up to appearance-based comments.
https://www.researchprotocols.org/2022/1/e33596 JMIR Res Protoc 2022 | vol. 11 | iss. 1 | e33596 | p. 4
(page number not for citation purposes)
XSL• FO
RenderXJMIR RESEARCH PROTOCOLS Garbett et al
Video Five: Be Your Own Best Friend! (targeting body talk)
Key messages:
• Negative body talk is harmful to our body image.
• Creating a mantra to repeat to oneself instead of getting caught up in negative self-talk is an effective strategy to break the cycle of negative body
talk.
Reinforcer activities:
• Create a mantra that can boost your confidence.
• List things that your body allows you to do.
• Your Own Words activity: Write about your mantra, sharing why you created it and what it means to you.
Video Six: The Color of the Future
Key messages:
• By challenging appearance pressures in everyday life, we can reach our full potential.
• The additive impact of resisting and challenging appearance pressures is large, not only for the individual but also for wider society.
Reinforcer activities:
• Watch a short video of Putri sharing the four key lessons she learned throughout her journey. Identify which lesson is most important to you and
why.
• Commit to sharing an unedited photo on social media doing something you love.
In addition to detailing the intervention and its development, as block randomization with a 1:1 allocation. Participants in the
this paper describes the evaluation protocol for assessing the intervention condition will receive the Warna-Warni Waktu
effectiveness of the intervention among young Indonesian intervention over a period of 6 days. Participants in the waitlist
women aged 15 to 19 years through a parallel, two-arm control condition will receive a link to the Warna-Warni Waktu
(intervention vs waitlist control) web-based randomized series at the end of the trial. The protocol was designed in
controlled trial (RCT). The decision to utilize a waitlist control accordance with the SPIRIT (Standard Protocol Items:
condition was informed by recommendations from a National Recommendations for Interventional Trials) guidelines
Institutes of Health expert panel [62]. Specifically, it is (Multimedia Appendix 1).
recommended that the rationale for a comparator group should
For the trial, the videos will not be distributed to recruited
rest on the primary purpose of the trial. Thus, given that the
participants via social media. Rather, the intervention will be
primary interest of this study was the absolute impact of the
recreated for distribution via Qualtrics software. Six Qualtrics
intervention rather than the relative impact, and that usual care
links will be developed, one for each video and its corresponding
for body dissatisfaction in Indonesia is no care at all, a waitlist
supplementary activities, and sent to participants daily over a
control condition was deemed most appropriate for this purpose.
6-day period. The trial has been designed in this way so that
Our hypotheses for this research are: (1) participants randomized
there is an accurate log of intervention adherence at the
to the intervention condition will experience reduced body
individual level, an important consideration when evaluating
dissatisfaction, internalization of appearance ideals, and skin
web-based interventions [63]. Careful consideration was given
shade dissatisfaction, as well as improvements in mood, at
to ensure that the intervention was accurately produced on
postintervention and 1-month follow-up, relative to the waitlist
Qualtrics to appear as closely as possible to how it will look
control condition; (2) each video in the Warna-Warni Waktu
when it is disseminated on social media in the future (see
series will elicit immediate state-based improvements in body
Multimedia Appendix 2).
satisfaction and mood; and (3) greater engagement and
adherence in the Warna-Warni Waktu intervention will result The Intervention
in greater reductions in body dissatisfaction, internalization of The intervention Warna-Warni Waktu was developed over a
appearance ideals, and skin shade dissatisfaction, as well as 20-month period, from October 2019 to May 2021. Textbox 2
greater improvements in mood. This analysis will be exploratory presents the intervention development steps taken, which
in nature, depending on the participants’ engagement and involved close collaboration among body image academics, a
adherence in the intervention during the trial. creative agency, social media specialists, a nonprofit
organization, as well as an industry funder. The intervention
Methods development team consisted of numerous women with lived
experience of Indonesian culture; namely, five of the six core
Study Design
team members from the creative agency (Percolate Galactic)
The study is a two-arm parallel RCT with an intervention group were young Indonesian women between the ages of 25 and 30.
and a waitlist control group. Randomization will be performed The wider team also consisted of two female Indonesian
https://www.researchprotocols.org/2022/1/e33596 JMIR Res Protoc 2022 | vol. 11 | iss. 1 | e33596 | p. 5
(page number not for citation purposes)
XSL• FO
RenderXJMIR RESEARCH PROTOCOLS Garbett et al
pediatricians (authors KN and BM), who have lived experience between 4 and 5 minutes each (see Figure 1) and 18 interactive
of the culture as well as daily contact with young Indonesian reinforcer activities, ranging from approximately 2 to 10 minutes
women through their work. Finally, the team included an each in length (see Figure 2 for examples). All content is
Indonesian Professor of Women’s Studies (LS). Further, the delivered to young women in Bahasa Indonesia, the official
process involved three rounds of feedback from young language of Indonesia. The activities encourage the target
Indonesian women at various stages of development to create audience to reflect and apply the learnings from the videos to
an intervention that is both for, and created with, young their own lives (Textbox 1).
Indonesian women. The final intervention comprises 6 videos
https://www.researchprotocols.org/2022/1/e33596 JMIR Res Protoc 2022 | vol. 11 | iss. 1 | e33596 | p. 6
(page number not for citation purposes)
XSL• FO
RenderXJMIR RESEARCH PROTOCOLS Garbett et al
Textbox 2. Intervention development stages.
Literature review: October 2019-January 2020
• Common appearance concerns among young Indonesian women include feeling pressure about their weight [64,65], skin shade [25], and skin
complexion [66].
• Prominent sources of appearance pressure contributing to negative body image among young Indonesian women were identified as cyberbullying
[67,68], appearance comparisons [66], and social media [66].
• eHealth interventions in low- and middle-income countries are increasing in number [47] and are acceptable among young people in Asia [69-72].
Secondary data analysis: December 2019-April 2020
• Secondary data analysis was conducted on data collected from 318 Indonesian girls and young women who participated in The Dove Global
Girls Beauty and Confidence Report [73], corroborating the literature review results and further identifying internalization of appearance ideals
and self-esteem as important influences of body image concerns (unpublished data of the authors KG, NC, LS, and PD, 2021).
U-Report poll on appearance-related concerns: February 2020
• In collaboration with UNICEF Indonesia, a U-Report poll was conducted among 1441 young women from all 34 Indonesian provinces for an
up-to-date assessment of the role body image plays in the lives of young people.
• More than three-quarters of young women wanted to change something about their appearance, and nearly half reported that worrying about their
appearance prevented them from doing things they would like to do. Nearly all young women reported that they would like to learn ways to
improve how they feel about their appearance [74].
Focus groups: March 2020
• Each of the six focus groups (one face-to-face and five online due to the COVID-19 pandemic) consisted of five or six girls aged 13-18 years
from Jakarta province.
• Appearance-based teasing and comments as well as pressure from social media (particularly influencers) were identified as prominent sources
of appearance pressure.
• Positive body image traits (such as valuing body functionality and defining beauty broadly, beyond physical appearance) were identified.
Intervention’s key messages: April-August 2020
• Four risk factors for the development of body image concerns were identified: (1) social media and influencers, (2) appearance-based comparisons,
(3) appearance-based teasing, and (4) body talk.
• It was decided to reinforce the positive body image traits throughout the intervention identified during the aforementioned focus groups.
Mode and format of delivery: April-August 2020
• Storytelling was chosen as the mode of information delivery, as narrative health education is proven to be effective [75] and engaging, even
among those with low health literacy [76]. (See Multimedia Appendix 3 for a synopsis of the intervention’s narrative).
• Videos were chosen as the intervention format to deliver the core messages owing to their efficiency in delivering content information (see Media
Richness Theory [77]) and ability to pique an audience’s attention and interest [78].
• Interactive activities to reinforce the key messages learned during the videos were added to elicit cognitive dissonance, an effective change
technique seen throughout the body image intervention literature [79,80].
Concept creation: August 2020-January 2021
• Two storyboard concepts were considered: time travel and a detective-based storyline. Time travel was selected given the ability of this concept
to convey the additive impact of body image concerns to young people.
• A combination of animated characters and real people was used, as cartoon characters in health intervention studies have been shown to be
acceptable to younger and older adolescents [81-83], as is the combined use of cartoon characters and real people [84].
• The plot and story details were drafted and refined over several collaborative discussions among the various project stakeholders.
• Body image change techniques (including those based on psychoeducation and media literacy [54]) were embedded within the video narrative.
• Six time travel–based videos (ie, one for each target risk factor, plus introductory and concluding videos) were drafted using storyboards and
basic animation.
U-Report poll on appearance-based teasing: January 2021
• A second U-Report poll with 240 young women was conducted to provide clarity as to how appearance-based teasing presents among young
people in Indonesia.
• Poll results provided direction on how to address appearance-based teasing, showing that it is prevalent both online and face-to-face, and that
teasing toward young women usually comes from other women, either from friends or family members [85].
https://www.researchprotocols.org/2022/1/e33596 JMIR Res Protoc 2022 | vol. 11 | iss. 1 | e33596 | p. 7
(page number not for citation purposes)
XSL• FO
RenderXJMIR RESEARCH PROTOCOLS Garbett et al
Cocreation of videos with the target audience: February 2021
• Findings on rough animated versions of the videos from four online focus groups with young women aged 15-19 years (N=16) showed strong
comprehension, acceptability, and enjoyment of the videos.
• The young women provided direction regarding appropriate terminology to aid comprehension within their age group.
Casting, scripts, props, sets, filming: February-March 2021
• Detailed scripts were written, casting auditions held, sets and props sourced, and decisions on futuristic details such as makeup and style were
decided.
• Accurate representation of the diverse Indonesian population through the actors and animated characters involved ensuring that various ethnicities,
skin shades, hair textures, distinct types of religious dress, body sizes, and regional accents were included.
• The videos were filmed. Numerous alternative versions were shot for scenes considered potentially problematic for comprehension so that options
were available to explore with young women, if necessary.
Development of activities: February-April 2021
• Activities were developed to allow users to practice the information learned and/or to reflect on their own cognitions and behaviors in light of
the video content.
• The team was guided by Girl Effect and Percolate’s prior experience of disseminating similar interactive content for other health campaigns, as
well as the body image scholars’ expertise in effective change techniques.
• Eliciting cognitive dissonance throughout the activities was prioritized, for example, by asking users to speak out against appearance ideals,
challenge body talk, and stand up to those bullying based on appearance.
Acceptability testing of the complete intervention: April 2021
• Six online focus groups of young women aged 15-19 years (N=36) were conducted to assess intervention acceptability and comprehension,
revealing strong acceptability, comprehension, and enjoyment of the videos and activities (see the Intervention Acceptability section).
• No issues with comprehension were identified, and thus no alternative scenes were required.
Final intervention edits: May 2021
• Based on focus group feedback, minor edits were made to the instructions for the activities, which focused on making the tone less formal and
enhancing comprehension.
Figure 1. The video thumbnails for each of the 6 episodes in Warna-Warni Waktu.
https://www.researchprotocols.org/2022/1/e33596 JMIR Res Protoc 2022 | vol. 11 | iss. 1 | e33596 | p. 8
(page number not for citation purposes)
XSL• FO
RenderXJMIR RESEARCH PROTOCOLS Garbett et al
Figure 2. Example activities. The activity on the left asks users to identify the differences between an edited and unedited photo (Video 2, Activity 3).
The activity on the right asks users to complete the comic strip with a comment about why appearance-based comparisons are unnecessary (Video 3,
Activity 2).
Although the majority of the activities are short, four activities 0-0.98) on a 5-point Likert-type scale ranging from 1 (hated the
are more in-depth, which consist of writing up to 250 words video) to 5 (loved the video). Only one respondent reported not
expressively and/or self-compassionately about what they fully understanding one of the videos (Video Five: Be Your
learned in the video, titled “Your Own Words.” These activities Own Best Friend); all other respondents reported full
were deemed important for participants to complete given the understanding of each video. This finding was corroborated
emerging evidence that expressive and compassion-based with qualitative feedback from young women, whereby many
writing are effective at reducing body dissatisfaction across a successfully restated the key messages from each video.
range of settings [86,87]. As such, all participants in the trial
Similar positive findings were found for each of the activities.
that complete these activities will be entered into a competition
Across all activities, likeability scores ranged from 4.42 to 4.89
to win Rp50,000 (approximately US $3.50) in mobile phone
(SD 0.32-0.90) on a 5-point Likert-type scale ranging from 1
credit. The responses to this activity will be judged by two
(hated the activity) to 5 (loved the activity). The activities were
authors (KN and BM). Judging will be based upon the best
understood by almost all the young women; 94.44%-100% of
response being (1) relevant, (2) of adequate length, (3) original,
respondents reported fully understanding each activity. Again,
(4) potentially inspirational to other young women, and (4)
comprehension was further identified through examination of
sounding sincere/honest. This approach reflects how the
the young women’s qualitative responses to the activities, with
intervention will be disseminated on social media, should the
the majority of participants responding in a way that exemplified
trial find it to be effective (ie, girls who engage with the
the key messages learned. The authors were particularly keen
intervention once disseminated will have the opportunity to
to understand the likelihood of the target audience to complete
submit their response to the writing task for the chance to win
the longer Your Own Words tasks. Findings showed that 95%
Rp50,000 in phone credit). This is a tried-and-tested method to
of respondents reported that if they had the time, they would
improve engagement in activities that Girl Effect has used in
be interested in participating in these tasks.
previous health campaigns on social media.
Based on the acceptability findings, minor edits were made to
Intervention Acceptability the reinforcer activity instructions for greater clarity and
In April 2021, the intervention underwent acceptability testing comprehension. In addition, upon suggestion from the young
with 36 young women in six focus groups. All groups watched women, less formal language was adopted to improve
the entire series; however, to minimize participant burden, three engagement. No changes were made to the design or format of
groups completed the activities associated with Videos 1-3 and the activities.
three groups completed the activities for Videos 4-6, providing
acceptability feedback on each. The videos scored high on
likeability: scores across the videos ranged from 4.44 to 5 (SD
https://www.researchprotocols.org/2022/1/e33596 JMIR Res Protoc 2022 | vol. 11 | iss. 1 | e33596 | p. 9
(page number not for citation purposes)
XSL• FO
RenderXJMIR RESEARCH PROTOCOLS Garbett et al
Study Setting Participant Recruitment and Procedure
The web-based trial will be coordinated by a research agency The aim is to recruit 1800 young women aged 15-19 years via
based in Jakarta, Indonesia. The aim is to recruit young women a local research agency’s recruitment panel (ie, a database
from across 10 of the largest cities in Indonesia: Balikpapan, containing contact details of those who have previously taken
Bandung, Jakarta, Makassar, Manado, Medan, Palembang, part in research conducted by the agency and have agreed to be
Pontianak, Semarang, and Surabaya. contacted with regard to future research), with the aim to recruit
an equal number of participants from each age (ie, 15, 16, 17,
Eligibility Criteria 18, and 19 years of age). The CONSORT (Consolidated
Inclusion criteria for participation include identifying as a young Standards of Reporting Trials) flowchart is provided in Figure
woman aged between 15 and 19 years (the target age group for 3. Women and men over the age of 40 years will be contacted
the intervention); having their own mobile phone (to ensure via telephone and screened for whether they have a daughter
participants receive the WhatsApp notifications regarding the within the eligible age range. If the respondent has more than
study); and accessing Facebook or Instagram daily (so that the one daughter in the age range, only one will be eligible to avoid
sample consists of those who are most likely to access and possible contagion effects. The daughter who is the best fit in
engage with the intervention when it is disseminated via social terms of reaching the age quota will be selected. If this does not
media channels in the future). Exclusion criteria include already distinguish which daughter is selected, the daughter with the
following the Girl Effect brand (Springster) on any social media birth date closest to the date of contact will be selected. If the
site or having ever accessed the Springster website prior to respondent does not have a daughter between 15 and 19 years
enrolment (to avoid contamination effects); and, if under 18 of age, the recruiter will enquire if they know another family
years of age, not having written consent from a parent or with a daughter of this age. If so, the recruiter will request the
guardian. telephone number of that family and contact them. Only one
phone number will be requested per call.
Figure 3. Participant flowchart.
https://www.researchprotocols.org/2022/1/e33596 JMIR Res Protoc 2022 | vol. 11 | iss. 1 | e33596 | p. 10
(page number not for citation purposes)
XSL• FO
RenderXJMIR RESEARCH PROTOCOLS Garbett et al
Should an eligible daughter be 15-17 years old, the recruiter requested to enter their PIN on the first page of the link.
will read the parental information sheet to the parent. Parents Participants will complete state measures of body satisfaction
will then be requested to provide verbal consent for their and mood before watching the first video in the Warna-Warni
daughter’s participation and verify their identity and daughter’s Waktu series. State measures of body satisfaction and mood
age. Parents will then answer questions relating to their will be asked again immediately after the first video. Next,
socioeconomic status on behalf of their daughter before the participants will be presented with the activity for the first video.
recruiter requests to speak to the daughter. If the daughter is not Before exiting the link, participants will be asked to report on
present, the recruiter will request a call back. The daughter will the strength of their internet connectivity while watching the
then be screened for eligibility and informed verbal assent video (see the Measures section below). This process is repeated
obtained. Following the call, and provided the daughter is on Days 4-8 for the remaining five videos (and associated
eligible and gives verbal assent, the parental information sheet activities). Participants will not be sent reminders to view or
will be sent to the parent via WhatsApp, with informed parental engage with the content in these links.
consent obtained once more, this time written, over WhatsApp.
On Day 9 of the study (again, in the morning), participants in
WhatsApp was chosen as it is the most used communication
both conditions will be sent a link to complete the second
app in Indonesia [88].
assessment. As with the baseline assessment, participants will
Should an eligible daughter be aged 18 or 19 years, a similar be given 24 hours to complete this assessment, with reminder
pattern of communication will occur. Parents will verify their messages sent to noncompleters after the first 8 hours. The same
own and their daughter’s identity and respond to questions process will be executed for the third and final assessment, 1
regarding the family’s socioeconomic status. Rather than parents month later on Day 36. Following the third assessment, all
providing informed verbal and written consent, this will be participants will be debriefed on the study aims and provided
completed by the daughters themselves, in the same manner with additional sources of mental health support, as well as a
with which it will be completed by parents of those aged 15-17 certificate of participation. Participants who complete all three
years. Verification of identities and ages will be achieved assessments will be rewarded with Rp125,000 (approximately
through video calls via the presentation of official US $8.75) to encourage participant retention. By this point, the
documentation (eg, National ID card, family registration card, responses to the four 250-word writing activities will have been
driving license, student ID). Recruitment is anticipated to take judged, and those who have won will be contacted and will
15 days (inclusive of weekends). receive their prize. Those in the waitlist control condition will
be provided with a link to the Warna-Warni Waktu video series
All participants will enter the study (ie, complete the baseline
on Day 37. Participants in the waitlist control condition will not
assessment survey) on the same day (Day 1). Participants will
be assessed for their engagement with the content. Information
receive a data package a day prior to this to ensure they have
sheets and the participant debrief document can be found in
ample mobile phone data to allow them to participate in the
Multimedia Appendix 4.
study. A link to the baseline assessment hosted on Qualtrics
(Provo, UT) will be sent to participants via WhatsApp at 8 AM Randomization and Blinding
(UTC +7) on Day 1, along with a unique participant Participants will be randomized following completion of the
identification number (PIN). Participants will be requested to baseline survey. Participants will not be blinded due to the nature
enter their PIN on the first page of the baseline survey, in order of the trial design. They will not be told explicitly of their
to match participant responses over time. Participants will have condition but will be made aware on Day 2 (ie, the day after
24 hours to complete the baseline assessment; those who have completing the baseline survey) when they will receive the
not completed the baseline assessment within the first 8 hours intervention. Randomization and allocation will be performed
will be sent a reminder message during the early evening of on an individual level using an automated, web-based
Day 1. Following the 24-hour window, participants who have randomizer to sequentially allocate participants based on a block
completed the baseline assessment will be randomized into one design to ensure a balance of participants across conditions. A
of two conditions: the intervention condition or a waitlist control researcher external to the project will generate the allocation
condition. Participants will be alerted on Day 2 to what happens sequence and assign participants to conditions. They will be
next, depending on which condition they have been randomized. blinded to participant information and conditions (ie, they will
Those randomized to the intervention condition will be informed be given a list of participant identification numbers and asked
that they can expect a series of sequential links to be sent to to assign them to Group A or B using the web-based
them daily over the following 6 days. Each link will contain randomizer). Data analysts will be blinded to condition
one episode of Warna-Warni Waktu and its associated activities. throughout the trial and analysis of the primary and secondary
The links will again direct participants to Qualtrics, where the trait-based measures. Blinding of data analysts during
intervention has been embedded (see Multimedia Appendix 2). state-based analyses is not possible due to the within-group
These participants will be requested to engage with the content design of this aspect of the trial. Two separate data files will be
of those links daily. Those in the control condition will be told provided to the statistician to ensure that blinding during
that they will be recontacted in 1 week to complete a second analyses of trait measures is retained. Due to the trial design,
assessment. the research agency and participants will not be blinded to
On the morning of the third day (Day 3), participants in the condition.
intervention condition will be sent their PIN and a link to the
first video (and associated activity). Again, participants will be
https://www.researchprotocols.org/2022/1/e33596 JMIR Res Protoc 2022 | vol. 11 | iss. 1 | e33596 | p. 11
(page number not for citation purposes)
XSL• FO
RenderXJMIR RESEARCH PROTOCOLS Garbett et al
Measures statistical consultants Chloe Hayes and Silia Vitoratou, 2021)
Self-report measures selected for use are presented in Table 1. were again selected for similar reasoning. Owing to the lack of
For hypothesis one, the primary outcome measure, the Body a validated assessment tool to assess skin shade satisfaction, a
Esteem Scale for Adults and Adolescents (BESAA) [89], was purpose-built measure will be used to assess this. For hypothesis
selected as it is the only body dissatisfaction measure to have two, single-item measures of body satisfaction and mood will
been validated among young people in Indonesia (unpublished be utilized. Full questionnaires are presented in Multimedia
data of the authors KG, NC, SH, KN, BM, LS, and PD, along Appendix 5.
with statistical consultants Chloe Hayes and Silia Vitoratou, Intervention adherence will be measured objectively through
2021) and evidence that this scale is amendable to change dwell time spent on the Qualtrics page that contains the video
following interventions of similar duration and content [90,91]. and via participant responses to the activities. Intervention
Two secondary outcome measures assessing internalization of acceptability will be assessed through six self-report items;
appearance ideals [92] and trait mood [93] (unpublished data these were informed by established acceptability frameworks
of the authors KG, NC, SH, KN, BM, LS, and PD, along with [96].
https://www.researchprotocols.org/2022/1/e33596 JMIR Res Protoc 2022 | vol. 11 | iss. 1 | e33596 | p. 12
(page number not for citation purposes)
XSL• FO
RenderXJMIR RESEARCH PROTOCOLS Garbett et al
Table 1. Measures.
Measure Description Time assessed
Demographics Age, country of birth, ethnicity, religion, socioeconomic status, social media usage T1a
Primary outcome: Trait body dissatisfaction Body Esteem Scale for Adolescents & Adults (BESAA) [89] adapted and validated T1, T2b, T3c
among adolescents in Indonesia (17 items). Example item: I like how I look like in
photos. Response options range from strongly disagree to strongly agree. Mean scores
range between 1 and 5, with higher scores reflecting higher body esteem.
Secondary outcomes
Internalization of appearance ideals Internalization-General subscale of the Sociocultural Attitudes Towards Appearance T1, T2, T3
Scale-3 [92] adapted and validated among adolescents in Indonesia (12 items). Ex-
ample item: I compare my body to the bodies of people who are on TV. Response
options range from strongly disagree to strongly agree. Mean scores range between
1 and 5, with higher scores reflecting higher internalization of appearance ideals.
Trait mood The Positive and Negative Affect Schedule for Children [93] adapted for Indonesia. T1, T2, T3
The validation of this scale among Indonesian adolescents is underway. The factor
structure will be determined prior to data analysis and reported in the main effective-
ness paper.
Skin shade satisfaction A purpose-built, single-item, skin shade discrepancy measure will assess skin shade T1, T2, T3
satisfaction. The chart comprises nine skin colors from dark (1) to light (9). Partici-
pants are asked to select the shade that most accurately matches their “current skin
shade” and the shade that most accurately reflects the skin shade they would prefer
(their “ideal skin shade”). The discrepancy between the two will be used as an indi-
cator of skin shade satisfaction, with higher absolute values indicating less skin shade
satisfaction. The score ranges from 0 (satisfied with skin shade) to 9 (very dissatisfied
with skin shade). The measure uses colors from The Pantone Skin Tone Guide [94].
State body satisfaction A single 101-point visual analog scale [95] will assess the immediate impact of each Before and after
video on participants’ state body satisfaction (ie, How satisfied are you with your each video
appearance, right now, in this moment?). The total score ranges from 0 (not at all)
to 101 (very much). Higher scores reflect greater satisfaction.
State mood A single 101-point visual analog scale [95] will assess the immediate impact of each Before and after
video on participants’ state mood (ie, How happy are you, right now, in this moment?). each video
The total score ranges from 0 (not at all) to 101 (very much). Higher scores reflect
greater positive mood.
Intervention adherence Digital metrics will assess participants’ engagement with the videos and activities, Six-day interven-
including: (1) percentage of participants who watch each video (determined by the tion period
number of participants whose dwell time on the Qualtrics page hosting each video
is equal to or longer than the length of the video); (2) percentage of participants who
watch all 6 videos (determined by the number of participants whose dwell time on
all 6 Qualtrics pages hosting videos is equal to or longer than the length of each
video); (3) average number of videos watched; (4) percentage of participants who
complete each activity; (5) average number of activities completed; and (6) average
length of time spent engaging with the intervention (ie, watching videos and complet-
ing activities) over the 6-day period.
Intervention acceptability Six items will assess participants’ acceptability of the intervention. Factors include T2
emotive response to the interventions (eg, enjoyment, likability of characters); rele-
vance (eg, age appropriateness, helpfulness, ease of understanding); ease of use (eg,
speed and accuracy of responses); and willingness to recommend (eg, how likely the
user would recommend the intervention to a friend). Item scores range between 1
(strongly disagree) and 5 (strongly agree). Higher scores reflect greater acceptability.
Only those randomized to the intervention condition will complete these measures.
a
T1: baseline assessment.
b
T2: postintervention assessment.
c
T3: follow-up assessment.
standardized effects sizes with Hedges g ranging from 0.25 to
Sample Size 0.4 (eg, [89]) with a standardized minimum important clinical
The predefined primary outcome measure is the BESAA at difference (MICD) of 0.2 exceeding the minimum detectable
postintervention assessment (T2) and follow-up assessment difference of the measures. To detect the MICD or larger, our
(T3). Similar RCTs assessing body dissatisfaction evaluating proposed sample size of n=900 per group would provide in
the same outcome measure reported a range of small to medium excess of 90% power (two-sided α=.05) for between-group
https://www.researchprotocols.org/2022/1/e33596 JMIR Res Protoc 2022 | vol. 11 | iss. 1 | e33596 | p. 13
(page number not for citation purposes)
XSL• FO
RenderXJMIR RESEARCH PROTOCOLS Garbett et al
differences at either T2 or T3 with or without a Bonferroni Ethics
correction for multiple time points. This assumes that dropout This study received ethical approval from the Faculty of
does not exceed 20% in any one arm, and is valid for any Medicine Universitas Indonesia
positive correlation between commensurate measures at baseline (588/UN2.F1/ETIK/PPM.00.002/2021) and the University of
(T1) with T2 or T1 and T3. The oversampling includes a the West of England (HAS.21.04.138). Participation in the study
COVID-19 contingency plan permitting the sample size to drop will be completely voluntary. For participants under 18 years
to 650 per group should external challenges arise, and further of age, parents will be approached first and will be required to
downward revisions in sample size with a tradeoff in power give their consent for their daughter’s participation. Prior to
under worst-case scenarios. consent, parents will be provided with a detailed information
Analysis Plan sheet outlining the requirements of participation, withdrawal
procedures, as well as potential risks. When parents provide
The intention-to-treat set of participants will form the primary
their consent, participant assent will be obtained, with the details
analysis set. An assessment of the impact of missing data on
of the study outlined again in a participant information sheet.
statistical conclusions will be undertaken using sensitivity
The same participant information sheet will be given to those
analyses.
participants aged 18 or 19 (without prior parental consent being
For the first hypothesis, the data will be analyzed on an sought). Participants will be informed that they can withdraw
intention-to-treat basis using a linear mixed model with baseline their consent at any point in the research process without
measures at T1 as a covariate, randomized group as a two-level needing to give justification. Parental consent and participant
between-subjects factor, and study phase (T2, T3) as two-level assent (or consent if 18 or 19 years old) will be obtained by the
repeated-measures factor. The statistical model will be research agency.
hierarchically balanced with the three-way interaction between
All information sheets will contain details of two counseling
covariate, phase, and randomized group as the generating class.
services available to young women in Indonesia if they are
This structure permits an analysis of covariance prior reasoned
experiencing any mental health concerns and require additional
comparison between randomized arms at T2 and T3, assessing
support. Further, information sheets will contain the contact
the parallel lines of assumption, homogeneity of variance
details of study author (BM) should they have any concerns
assumption, and the use of robust estimates, if needed. Changes
relating to the execution of the study. The study is registered
within the randomized arm between T1 and T2 and between T1
with ClinicalTrials.Gov (NCT05023213).
and T3 will be assessed using the paired-samples t-test and
effect size quantified with 95% CIs. The reliable change index Special ethical consideration was given in light of conducting
will be used to determine the percentage of participants reliably this research during the COVID-19 pandemic. Although at the
improving within each arm. time of writing, legal regulations would allow for face-to-face
contact between the research agency and participants, it was
For the second hypothesis, the nested intervention study
decided that the research should take place entirely online.
comprises state evaluation of body satisfaction and mood
Although such a strategy is befitting for the evaluation of a
immediately before and after each daily video. A component
web-based intervention, assurances regarding identity during
impact analysis will consist of a 2 (pre, post) by 6 (Day 3 to
the recruitment phase required careful consideration. Video
Day 8) fully repeated-measures analysis including linear and
calls where parents of potential participants will be required to
quadratic trends for time sequence effects, and a main-effects
show official photographic identification will allow for
comparison between components on pre and post change scores.
confirmation of participant identities, including age. Owing to
It is not inconceivable that relative effects between days are
the web-based nature of the study, the decision to only include
likely to be small. For an assumed standardized effect of Cohen
participants who had access to their own mobile phone was
d=0.1, a sample size of n=800 would be needed for 80% power.
made, ensuring that only the recruited participants receive, and
These data have further value in evaluating adherence and would
respond to, the notifications sent about the study. Anecdotally,
permit a dose-response effect to be included in a planned subset
the Indonesian-based authors believe that many young women
analysis.
in the target age group own their own mobile phone; therefore,
For the third hypothesis, the relationship between adherence we do not anticipate this impacting the representativeness of
(count of daily completion in each of Day 3 to Day 8) and the sample obtained. For transparency, participants ineligible
outcome (primary and secondary) at each of T2 and T3 will be due to not owning their own mobile phone will be reported in
assessed using linear regression controlling for baseline the main trial.
covariates. The cardinal nature of adherence permits Helmert
effect coding (difference effect coding) to be used to estimate
Data Monitoring and Management
cumulative dosage effects. The relationship between daily Data from recruited participants will be downloaded directly
engagement and outcome (primary and secondary) at T2 and from Qualtrics; thus, there is no data entry process to consider.
T3 will be assessed using linear regression controlling for Once downloaded, the data will be confirmed correct by
baseline covariates. These latter models will code each daily checking that the data values are as expected. Participant
engagement as a dummy variable, and will permit a comparison responses will be matched over time, and any duplication of
between engagement in each activity. A full statistical analysis responses will be examined, and if found, deleted. No personal
plan will be written and approved by the Trial Management details will be requested from participants via Qualtrics, and as
Group prior to study closure. such, these data files should be anonymous. However, the files
https://www.researchprotocols.org/2022/1/e33596 JMIR Res Protoc 2022 | vol. 11 | iss. 1 | e33596 | p. 14
(page number not for citation purposes)
XSL• FO
RenderXYou can also read
